Materials and methods
Virus and cell culture
African green monkey kidney fibroblast CV-1 cells and human pancreatic ductal carcinoma PANC-1 cells were purchased from American Type Culture Collection (ATCC) (Manassas, VA) and were grown in Dulbecco's modified Eagle's medium (DMEM) supplemented with 1% antibiotic-antimycotic solution (Mediatech, Inc., Herndon, VA) and 10% fetal bovine serum (FBS) (Mediatech, Inc.) at 37°C under 5% CO
2. Rat thyroid PCCL3 cells were a kind gift from the lab of Dr. James Fagin at MSKCC and were maintained in Coon's modified medium (Sigma, St. Louis, MO), 5% calf serum, 2 mM glutamine, 1% penicillin/streptomycin, 10 mM NaHCO3, and 6H hormone (1 mU/ml bovine TSH, 10 ug/ml bovine insulin, 10 nM hydrocortisone, 5 ug/ml transferrin, 10 ng/ml somatostatin, and 2 ng/ml L-glycyl-histidyl-lysine) at 37°C under 5% CO
2. GLV-1h68 was derived from VACV LIVP, as described previously [
6].
Construction of hNIS transfer vector
The hNIS cDNA was amplified by polymerase chain reaction (PCR) using human cDNA clone TC124097 (SLC5A5) from OriGene as the template with primers hNIS-5 (5'-GTCGAC(Sal I) CACCATGGAGGCCGTGGAGACCGG-3') and hNIS-3 (5'-TTAATTAA(Pac I) TCAGAGGTTTGTCTCCTGCTGGTCTCGA-3'). The PCR product was gel purified, and cloned into the pCR-Blunt II-TOPO vector using Zero Blunt TOPO PCR Cloning Kit (Invitrogen, Carlsbad, California). The resulting construct pCRII-hNIS-1 was sequenced, and found to contain an extra 33-bp segment in the middle of the coding sequence, representing an alternative splicing product for hNIS. To remove this extra 33-bp segment, two additional primers were designed to flank the segment, and used in the next set of PCR. In the next round of reactions, hNIS-5 paired with hNIS-a3 (5'-GAGGCATGTACTGGTCTGGGGCAGAGATGC-3'), and hNIS-a5 (5'-CCCAGACCAGTACATGCCTCTGCTGGTGCTG-3') paired with hNIS-3 were used in separate PCRs, both with pCRII-hNIS-1 as the template. The respective PCR products were then mixed and used as the templates in one reaction with hNIS-5 and hNIS-3 as the primer pair. The final PCR product was again cloned into the pCR-Blunt II-TOPO vector as pCRII-hNISa-2, confirmed by sequencing to be identical to the SLC5A5 sequence in GenBank (accession number NM_000453). The hNIS cDNA was then released from pCRII-hNIS-1 with Sal I and Pac I, and subcloned into HA-SE-RLN-7 with the same cuts by replacing RLN cDNA. The resulting construct HA-SE-hNIS-1 were confirmed by sequencing and used for insertion of PE-hNIS into the HA locus of GLV-1h68.
Generation of hNIS-expressing VACV
CV-1 cells were infected with GLV-1h68 at a multiplicity of infection (MOI) of 0.1 for 1 hour, then transfected using Fugene (Roche, Indianapolis, IN) with the hNIS transfer vector. Two days post infection, infected/transfected cells were harvested and the recombinant viruses selected and plaque purified as described previously [
14]. The genotype of hNIS-expressing VACV GLV-1h153 was verified by PCR and sequencing. Also, expression of GFP and β-galactosidase was confirmed by fluorescence microscopy and 5-bromo-4-chloro-3-indolyl-β-D-galactopyranoside (X-gal, Stratagene, La Jolla, CA), respectively, and lack of expression of
gusA was confirmed by 5-bromo-4-chloro-3-indolyl-β-D-glucuronic acid (X-GlcA, Research Product International Corp., Mt. Prospect, IL).
Viral growth curves
PANC-1 cells were seeded onto 6-well plates at 5 × 105 cells per well. After 24 hours in culture, cells were infected with either GLV-1h153 or GLV-1h68 at an MOI of 0.01 or 1.0. Cells were incubated at 37°C for 1 hour with brief agitation every 30 minutes to allow infection to occur. The infection medium was then removed, and cells were incubated in fresh growth medium until cell harvest at 1, 24, 48, and 72 hours post infection. Viral particles from the infected cells were released by 3 freeze-thaw cycles, and the titers determined as (PFU/106) in duplicate by plaque assay in CV-1 cell monolayers.
Flow cytometry
Cells were seeded on 6-well plates at 5 × 105 cells per well. Wells were then infected at MOIs of 0, 0.01, and 1.0, and cells then harvested at 6, 12, 24, 48, 72, and 96 hours postinfection by trypsinizing and washing with phosphate-buffered saline (PBS). For the second experiments, cells were seeded on 6-well plates at 5 × 105 cells per well. Wells were then infected at MOIs of 0, 0.01, 0.1, 0.5, 1.0, 2.0, and 5, and were harvested in the same manner at 24 hours after infection. GFP expression was analyzed via a Becton-Dickinson FACScan Plus cytometer (Becton-Dickinson, San Jose, CA). Analysis was performed using CellQuest software (Becton-Dickinson).
hNIS mRNA analysis via microarray
To evaluate the level of hNIS mRNA production in infected cells, cells were plated at 5 × 105 cells per well and infected with GLV-1h153 at an MOI of 5.0. Six and 24 hours postinfection, 3 samples of each time point were harvested and lysis performed directly using RNeasy mini kit protocol (Qiagen Inc., Valencia, CA). The mRNA samples were measured by spectrophotometer for proof of purity and hybridized to HG-U133A cDNA microarray chips (Affymetrix Inc, Santa Clara, CA) by the genomic core laboratory at Memorial Sloan-Kettering Cancer Center (MSKCC). The chip images were scanned and processed to CEL files using the standard GCOS analysis suite (Affymetrix Inc). The CEL files were then normalized and processed to signal intensities using the gcRMA algorithm from the Bioconductor library for the R statistical programming system. All subsequent analysis was done on the log (base 2) transformed data. To find differentially expressed genes a moderated t-test was used as implemented in the Bioconductor LIMMA package. To control for multiple testing the False Discovery Rate (FDR) method was used with a cutoff of 0.05.
hNIS protein analysis via Western blot
To confirm whether the hNIS protein was being expressed in infected cells, cells were plated at 5 × 105 per well and infected with GLV-1h153 at various MOIs of virus, harvested at 24 hours, and suspended with SDS-PAGE and 0.5-m DDT reagent. After sonication, 30 ug of the protein samples were loaded on 10% Bis-Tris-HCl buffered polyacrylamide gels using the Bio-rad system (Bio-rad laboratories, San Francisco, CA). Following gel electrophoresis for 1 hour, proteins were transferred to nitrocellulose membranes using electroblotting. Membranes were then preincubated for 1 hour in 5% low fat dried milk in TBS-T (20 mm Tris, 137 mm NaCl, and 0.1% Tween-20) to block nonspecific binding sites. Membranes were incubated with a purified mouse antibody against hNIS at 1:100 dilution (Abcam Inc., Cambridge, MA) and incubated for 12 hour at +4°C. After washing with TBS-T, secondary antibody (horseradish peroxidase-conjugated goat antimouse IgG (Santa Cruz, Santa Cruz, California) was applied for 1 hour at room temperature at a 1:5,000 dilution. Peroxidase-bound protein bands were visualized using enhanced chemiluminescence Western blotting detection reagents (Amersham, Arlington Heights, IL) at room temperature for approximately 1 minute and using Kodak BIOMAX MR films for exposure. Normal human thyroid lysate was used as a positive control, and cells treated with GLV-1h68 and PBS were used as negative controls.
Immunofluorescence
PANC-1 cells grown in a 12-well plate at 1 × 106 were mock-infected with GLV-1h68 or infected with GLV-1h153 at an MOI of 1.0. Twenty-four hours after infection the cells were fixed with 3.7% paraformaldehyde, permeabilized with methanol, blocked with PBS containing BSA, and incubated with a mouse anti-hNIS monoclonal antibody (Abcam Inc., Cambridge, MA) at a dilution of 1:100, followed by incubation with a secondary red fluorochrome-conjugated goat antimouse antibody (Invitrogen) at a dilution of 1:100. Pictures were taken using a Nikon inverted fluorescence microscope.
In vitro radiouptake assay
Radio-uptake in cells infected with GLV-1h153 was compared to rat thyroid cell line PCCL3 endogenously expressing NIS and to cells infected with parental virus GLV-1h68. Cells were plated at 5 × 105 per well in 6-well plates. Twenty-four hours after infection with MOIs of 0.01, 0.10, and 1.0, cells were treated with 0.5 μCi of either carrier-free 131I or 131I with 1 mM of sodium perchlorate (NaClO4), a competitive inhibitor of hNIS, for a 60-minute incubation period. Media was supplemented with 10 μM of sodium iodide (NaI). Iodide uptake was terminated by removing the medium and washing cells twice with PBS. Finally, cells were solubilized in lysis buffer for residual radioactivity, and the cell pellet-to-medium activity ratio (cpm/g of pellet versus cpm/mL of medium) calculated from the radioactivity measurements assayed in a Packard γ-counter (Perkin Elmer, Waltham, MA). Results were expressed as change in uptake relative to negative uninfected control. All samples were done in triplicate.
In vitro cytotoxicity assay
PANC-1 pancreatic cancer cells were plated at 2 × 104 per well in 6-well plates. After incubation for 6 hours, cells were infected with GLV-1h153 or GLV-1h68 at MOIs of 1.00, 0.10, 0.01, and 0 (control wells). Viral cytotoxicity was measured on day 1 and every second day thereafter by lactate dehydrogenase (LDH) release assay. Results are expressed as the percentage of surviving cells as compared to uninfected control.
In vivo tumor therapy studies and systemic toxicity
All mice were cared for and maintained in accordance with animal welfare regulations under an approved protocol by the Institutional Animal Care and Use Committee at the San Diego Science Center, San Diego, California. PANC-1 xenografts were developed in 6- to 8-week-old male nude mice (NCI:Hsd:Athymic Nude-Foxn1 nu, Harlan) by implanting 2 × 106 PANC-1 cells in PBS subcutaneously in the left hindleg. Tumor growth was recorded once a week in 3 dimensions using a digital caliper and reported in mm3 using the formula (length × width × [height-5]). When tumors reached 100-300 mm3, mice were injected intratumorally (IT) or intravenously (IV) via the tail vein with a single dose of 2 × 106 PFUs of GLV-1h153 or GLV-1h68 in 100 μL PBS. Animals were observed daily for any sign of toxicity, and body weight checked weekly.
Radiopharmaceuticals
124I and 131I were obtained from MSKCC's radiopharmacy. The maximum specific activities for the 124I and 131I compounds were ~140 μCi/mouse and ~0.5 μCi/well, respectively.
In vivo PET imaging
All animal studies were performed in compliance with all applicable policies, procedures, and regulatory requirements of the Institutional Animal Care and Use Committee, the Research Animal Resource Center of MSKCC, and the National Institutes of Health "Guide for the Care and Use of Laboratory Animals." Three groups of 2-3 animals bearing subcutaneous PANC-1 xenografts on the left hindleg measuring were injected intratumorally with 2 × 107 PFU GLV-1h153 (3 mice), 2 × 107 PFU GLV-1h68 (2 mice), or PBS (2 mice). Two days after viral injection, 140 μCi of 124I was administered via the tail vein. One hour after radiotracer administration, 3-dimensional list-mode data were acquired using an energy window of 350 to 700 keV, and a coincidence timing window of 6 nanoseconds. Imaging was performed using a Focus 120 microPET dedicated small animal PET scanner (Concorde Microsystems Inc, Knoxville, TN). These data were then sorted into 2-dimensional histograms by Fourier rebinning. The image data were corrected for (a) nonuniformity of scanner response using a uniform cylinder source-based normalization, (b) dead time count losses using a single-count rate-based global correction, (c) physical decay to the time of injection, and (d) the 124I branching ratio. The count rates in the reconstructed images were converted to activity concentration (%ID/g) using a system calibration factor (MBq/mL per cps/voxel) derived from imaging of a mouse-size phantom filled with a uniform aqueous solution of 18F. Image analysis was performed using ASIPro (Siemens Pre-clinical Solutions, Knoxville, TN).
Statistical analysis
The GraphPad Prism 5.0 program (GraphPad Software, San Diego, CA) was used for data handling and analysis. The significance of differences between the 3 therapy groups (untreated, GLV-1h153, GLV-1h68) was determined via two-way ANOVA with Bonferroni correction. P values were generated for radiouptake assay comparisons using Dunnett's test [
15]. P < 0.05 was considered significant.
Discussion
Oncolytic viral therapy is emerging as a novel cancer therapy. Preclinical and clinical studies have shown a number of oncolytic viruses to have a broad spectrum of anti-cancer activity and safety [
18]. These are ongoing, and the first oncolytic viral therapy has now been approved in China as a treatment for head and neck cancers [
19]. Clinical trials are underway to assess the effects of many other oncolytic viral therapies [
20]. However, future clinical studies may benefit from the ability to noninvasively and serially identify sites of viral targeting and to measure the level of viral infection and spread in order to provide important information for correlation with safety, efficacy, and toxicity [
3‐
5]. Such real-time tracking would also provide useful information regarding timing of viral dose and administration for optimization of therapy, as well as distribution and replication of the oncolytic virus, and would alleviate the need for multiple and repeated tissue biopsies.
VACV is arguably the most successful biologic therapy agent, since versions of this virus were given to millions of humans during the smallpox eradication campaign [
2]. More recently, engineered VACVs have also been successfully used as direct oncolytic agents, capable of preferentially infecting, replicating within, and killing a wide variety of cancer cell types [
6‐
11,
13,
21]. VACV displays many of the qualities thought necessary for an effective oncolytic antitumor agent. In particular, the large insertional cloning capacity allows for the inclusion of several functional and therapeutic transgenes. With the insertion of reporter genes not expressed in uninfected cells, viruses can be localized and the course of viral therapy monitored in patients.
One such promising virus strain is GLV-1h68[
21]. This strain has shown efficacy in the treatment of a wide range of human cancers and is currently being tested in phase I human trials[
20]. In this study we describe the generation of a novel recombinant VACV, GLV-1h153, derived from GLV-1h68, which has been engineered for specific targeted treatment of cancer and the additional capability of facilitating noninvasive imaging of tumors and metastases. To our knowledge, GLV-1h153 is the first oncolytic VACV expressing the hNIS protein.
The reporter gene chosen for insertion into GLV-1h153 was based on the already successful PET and SPECT imaging characteristics of the human sodium iodide symporter (hNIS) and carrier-free radioiodine reporter imaging system. hNIS is an intrinsic plasma membrane protein that mediates the active transport and concentration of iodide in thyroid gland cells and some extra-thyroidal tissues [
16,
17]. Although endogenous NIS is physiologically and functionally expressed in several normal tissues, so far only 2 human cancers - some thyroid cancers, and around 80% of breast cancers including ductal carcinomas - have been shown to express endogenous NIS functionally, making them amenable to radiotherapy [
22]. It is one of several human reporter genes that are currently being used in preclinical studies and has even been used in clinical studies imaging prostate cancer [
20,
23]. hNIS gene transfer via viral vector may allow infected tumor cells to concentrate several commercially available, relatively inexpensive radionuclide probes, such as
123I,
124I,
125I, and
99mTcO
4, all of which have long been approved for human use by the U.S. Food and Drug Administration, allowing noninvasive imaging of tumors expressing NIS [
22]. In contrast to a study published by McCart
et al.[
24] using an oncolytic VACV expressing the human somatostatin receptor hSSTR2, hNIS is a transporter-based reporter gene system. Whereas receptors usually have a 1:1 binding relationship with a radiolabeled ligand, transporters provide signal amplification through transport-mediated concentrative intracellular accumulation of substrate. hNIS use has also been shown to be comparable to the commonly used
HSV1-tk reporter gene [
25] and correlated with
99mTcO
4[
26]. This can be very useful for viral distribution with scintigraphy or PET scanning during and after viral therapy, and may allow for correlation with efficacy and toxicity during clinical trials and treatment thus offering potential clinical translation of this dual therapy.
In order to take advantage of the therapeutic and imaging potential of hNIS, several groups have attempted exogenous
NIS gene transfer in several human cancers including head and neck squamous cell cancers, non-small cell lung, thyroid, liver, colorectal, and prostate cancers, as well as glioma and multiple myeloma [
22]. Studies have shown that
hNIS gene delivery to both thyroidal and non-thyroidal, non-organifying tumor cells is capable of inducing accumulation of therapeutically effective radioiodine doses. For example, a single therapeutic
131I dose of 3 mCi was shown to elicit a dramatic therapeutic response in NIS-transfected prostate cell xenografts, with an average volume reduction of more than 90% [
27]. Transfection of an
hNIS-defective follicular thyroid carcinoma cell line with the
hNIS gene was able to reestablish iodide accumulation activity both in cell culture and in animal models [
28]. Furthermore, transfection of pancreatic cancer cells with a replication-deficient adenoviral vector expressing hNIS lead to a more than 15-fold increase in iodide uptake visualized with
123I scintigraphy, and an over 75% reduction in volume
in vivo after treatment with 3mci of
131I [
29].
We have previously reported on the use of a novel recombinant VACV, GLV-1h99, a derivative of GLV-1h68, which was constructed to carry the human norepinephrine transporter gene (hNET) under the VACV synthetic early promoter placed at the
F14.5L locus for deep tissue imaging [
30,
31]. The parental virus GLV-1h68, a recombinant VACV (LIVP strain), was constructed by inserting 3 expression cassettes (
Renilla luciferase-
Aequorea green fluorescent protein (RUC-GFP) fusion, β-galactosidase, and β-glucuronidase) into the
F14.5L,
J2R, and
A56R loci of the viral genome, respectively [
6]. The hNET protein was expressed at high levels on the membranes of cells infected with GLV-1h99, and expression of the hNET protein did not negatively affect virus replication in cell culture or
in vivo virotherapeutic efficacy. GLV-1h99-mediated expression of the hNET protein in infected cells resulted in specific uptake of the radiotracer [
131I]-meta-iodobenzylguanidine ([
131I]-MIBG). In mice, GLV-1h99-infected tumors, including pancreatic and mesothelioma, were readily imaged by [
124I]-MIBG PET. However, one of the disadvantages of using hNET is that it requires the carrier MIBG for radioiodine uptake.
GLV-1h153 was effective at infecting and replicating within pancreatic cancer PANC-1 cells as efficiently as its parental virus GLV-1h68. This indicated that insertion of the hNIS protein did not negatively affect virus replication in cell culture. With the hNET-expressing GLV-1h99 virus, on the other hand, there was slight improvement in viral replication and oncolytic effect, which may have been due to the exchange of expression cassettes under the control of promoters with different strengths.
Microarray analysis revealed an almost 2000-fold change increase in hNIS mRNA and an almost 5000-fold change increase by 24 hours after PANC-1 infection with GLV-1h153 at a multiplicity of infection of 5.0. Western blot studies showed hNIS protein expression as a band between 75 and 100 KiloDalton in PANC-1 cells infected with GLV-1h153, with higher concentrations of the protein at higher MOIs. This band also appears in normal human thyroid lysates at a slightly lower molecular weight, which is likely explained by differences in glycosylation within cells [
16]. The hNIS protein was successfully transported and inserted into the cell membrane, as demonstrated by fluorescence microscopy.
In vitro, GLV-1h153-mediated expression of the hNIS protein in infected PANC-1 cells resulted in specific uptake of the radiotracer
131I, indicating that the hNIS protein was functional, with the uptake reaching a >70-fold increase compared with uninfected control at an MOI of 1.0.
GLV-1h153 was also effective at infecting, replicating within, and killing PANC-1 cells and eradicating tumor xenografts as efficiently as parental virus GLV-1h68. This indicated that insertion of the hNIS protein did not negatively affect virus replication in vivo which was already demonstrated in vitro, or the cytolytic activity in cell culture and in vivo virotherapeutic efficacy. Similar effects were seen between the IT and IV groups treated with GLV-1h153 or GLV-1h68, indicating the inherent affinity of both genetically modified vaccinia viruses to tumors. Furthermore, administration of GLV-1h153 did not have any significant effects on mean net body weights of the animals 34 days after treatment, with the IT groups even gaining weight as compared to untreated control.
Finally, in mice, three PANC-1 tumors infected with GLV-1h153 were readily detectable by PET, with no enhancement above background of either GLV-1h68- or PBS-treated tumors. Mice were treated intratumorally with GLV-1h153, non-hNIS expressing parent virus GLV-1h68, and PBS, and imaged 48 hours after with carrier free 124I. The quantitative 124I -PET showed that imaging of GLV-1h153 infection of PANC-1 tumors is feasible after direct tumor injection.
As with any translational therapy, concerns over immune responses remain. Since
hNIS is a human derived gene, it is unlikely to be immunogenic. However, application of GLV-1h153-mediated hNIS transfer raises concerns over the possibility of autoimmunity in patients. Several papers have already shown that hNIS is not a major candidate for autoimmune disease in patients with patients with Graves' disease and Hashimoto's thyroiditis [
32,
33]. Moreover, a clinical trial assessing adenoviral-mediated hNIS transfer in humans did not report any serious adverse effects due to autoimmunity in patients treated for prostate cancer [
23]. Further studies and caution will be needed to assess the potential of autoimmunity with hNIS transfer in humans. Studies are also now underway to determine the viral biodistribution, as well as the dose and timing related aspects of
in vivo imaging using this novel virus.
Authors' contributions
NC was instrumental in the construction and homologous recombination of GLV-1h153. QZ was instrumental in the hNIS transfer construction and viral sequencing. CC contributed to study design, western blot, and immunofluorescence. YY contributed to the study design and assisted in in vivo studies. SC contributed to the cytotoxicity and flow cytometry assays. JC contributed to the in vitro radiouptake assays. JA contributed to the in vitro radiouptake assays. AM contributed to protein harvesting and western blot. MG was instrumental in statistical analysis of data. PZ was instrumental in study design, imaging experiments, and advice regarding hNIS biology and physiology. YF is the co-corresponding author and was critical to study design and completion. AZ is the corresponding author of this paper and was critical to study design and completion. All authors have read and approved the final manuscript.